Summary of the invention
The application's technical matters to be solved provides a kind of group buying websites sales data authenticity detection method and device, can solve the problem that group buying websites sales data authenticity is judged.
In order to address the above problem, the application discloses a kind of group buying websites sales data authenticity detection method, may further comprise the steps:
Obtain the real-time sales volume that shows in the single information of current group;
Obtain the single real-time attention rate of current group;
Obtain the real-time sales volume that shows in the single information of current group in the same period and the difference of the weighted value of attention rate in real time, if difference, confirms then that the single sales data of current group is untrue greater than threshold value, otherwise, confirm that then the single sales data of current group is true.
Further, the said real-time sales volume that shows in the single information of current group of obtaining comprises:
, the single real-time sales volume of current group obtains when reaching predetermined quantity.
Further, the single real-time attention rate of said current group is confirmed in the following way:
Obtain the number of times that current single-page clicked;
The said number of times correction of being clicked is obtained the single number of times of perhaps being browsed by user's normal access of said current group;
The said number of times of perhaps being browsed by user's normal access is the single real-time attention rate of current group.
Further, the said number of times correction of being clicked is comprised:
Visitor's ID is discerned and filtered.
Further; The weighted value of said real-time attention rate is handled for real-time attention rate being carried out normalization; Make it to have the normalization processing costs that obtains behind the identical quantized level with real-time sales volume, the real-time sales volume that shows in the single information of said current group with the difference of the weighted value of real-time attention rate is: the difference of the changing value of real-time sales volume changing value and normalization processing costs.
Further; The weighted value of said real-time attention rate is for transforming the theoretical quantity purchase that obtains with real-time attention rate, and the real-time sales volume that shows in the single information of said current group with the difference of the weighted value of real-time attention rate is: the difference of real-time sales volume and theoretical quantity purchase.
Further, said real-time attention rate is transformed obtains theoretical quantity purchase and comprises:
Confirm the single concrete commodity of current group according to the single information of single group of current group;
Confirm the purchase conversion ratio of said concrete commodity;
Confirm theoretical quantity purchase according to buying conversion ratio with real-time attention rate.
Further, the real-time sales volume that shows in the single information of current group in the said same period comprises with the difference of the weighted value of real-time attention rate:
The real-time sales volume of same time point and the difference of theoretical quantity purchase; Or
The real-time sales volume of each identical time point and the difference sum of theoretical quantity purchase in the same period.
In order to address the above problem, disclosed herein as well is a kind of group buying websites sales data authenticity detection method, may further comprise the steps:
Obtain the single information of single group of current group and roll into a ball the real-time sales volume that shows in single information;
Confirm that based on the single information of single group of current group single relevant group of said current group is single;
Obtain the single in real time total attention rate of single and relevant group of current group;
Obtain the difference of the weighted value of single and relevant the single in real time total attention rate of real-time sales volume and said current group that shows in the single information of current group in the same period; If difference is greater than threshold value; Confirm that then the single sales data of current group is untrue; Otherwise, confirm that then the single sales data of current group is true.
Further, the said single information of group based on current group list confirms that single relevant group of said current group singly comprises:
From the single information of single group of current group, extract the single keyword of current group;
The keyword set that is expanded expanded in keyword to extracting;
According to purchase by group cloud data inquiry and the relevant group list definite and expanded keyword coupling of etendue critical set of words from prestoring.
Further, saidly from the single information of single group of current group, extract the single keyword of current group and comprise:
Extract keyword from current the group single title and the detailed description.
Further, the single in real time total attention rate of single and relevant group of said current group is confirmed in the following way:
Obtain the total degree that single and relevant group of current group single-page is clicked;
The said total degree correction of being clicked is obtained the single total degree of perhaps being browsed by user's normal access of single and relevant group of said current group;
The said total degree of perhaps being browsed by user's normal access is the single real-time total attention rate of single and relevant group of current group.
In order to address the above problem, disclosed herein as well is a kind of group buying websites sales data authenticity pick-up unit, comprising:
The single information acquisition module of current group is used for obtaining the real-time sales volume that the single information of current group shows;
The attention rate determination module is used to obtain the single real-time attention rate of current group in real time;
Comparison module; Be used for the single information of current group shows in the more same period the real-time sales volume and the difference of the weighted value of attention rate in real time,, confirm that then the single sales data of current group is untrue if greater than threshold value; Otherwise, confirm that then the single sales data of current group is true.
Further, said device also comprises:
Trigger module is used for when the real-time sales volume that the single information of current group shows reaches predetermined quantity, triggering the real-time sales volume of obtaining current group list.
Further, said real-time attention rate determination module comprises:
The number of clicks acquiring unit is used to obtain the number of times that current single-page clicked;
Amending unit is used for the said number of times correction of being clicked is obtained the single number of times of perhaps being browsed by user's normal access of said current group;
The attention rate determining unit is used for confirming that by the number of times that user's normal access is perhaps browsed be the single real-time attention rate of current group in real time.
Further, said amending unit comprises:
ID recognin unit is used for visitor's ID is discerned and filtered.
Further, said device also comprises:
The single determination module of relevant group is used for confirming that based on the single information of single group of current group single relevant group of current group is single.
Further, the single determination module of said relevant group comprises:
Keyword extracting unit is used for extracting the single keyword of current group from the single information of single group of current group;
Expanding element is used for the keyword set that is expanded expanded in the keyword that extracts;
Query unit is used for purchasing by group the inquiry of cloud data and confirming single with the relevant group of expanded keyword coupling from what prestore based on the etendue critical set of words.
Further, said device also comprises:
Conversion module is used for real-time attention rate is converted into theoretical quantity purchase.
Further, said conversion module comprises:
The keyword acquiring unit is used to obtain the single keyword of current group;
Commodity are confirmed the unit, are used for confirming the single concrete commodity that purchased by group of current group based on the single keyword of current group;
Conversion ratio is confirmed the unit, is used for confirming the purchase conversion ratio of said concrete commodity;
Transformation model is used for confirming theoretical quantity purchase according to buying conversion ratio with real-time attention rate.
Compared with prior art, the application has the following advantages:
The application is through the real-time attention rate of the group's of obtaining list; And carry out weighted according to the data of statistics in advance and make it to passing judgment on a reference value of sales volume; Then through obtaining the single real-time sales volume of current group; Carry out difference relatively with the weighted value of real-time attention rate, thereby realize purchasing by group the detection of sales data authenticity.
Further; It is single to obtain its relevant group after the single information of group is analyzed; And pass through the real time data monitoring of relevant group list and the analysis of various particular commodity historical datas, current real-time attention rate of rolling into a ball single and relevant group list is converted into a relatively objective real theoretical quantity purchase, through the real-time sales data of more current group list and the difference of theoretical quantity purchase; Thereby judge whether the single real-time sales data of current group is true, thereby make judged result more accurate.
In addition; Transforming real-time attention rate is in the process of theoretical quantity purchase; Through in the server in advance the various group buying websites of storage purchase by group the processing that the cloud data are carried out related data, need not to reanalyse and obtain, this just makes the foundation of conversion more objective; The process of handling is also more simple, can guarantee the accuracy of final judged result simultaneously.
Embodiment
For above-mentioned purpose, the feature and advantage that make the application can be more obviously understandable, the application is done further detailed explanation below in conjunction with accompanying drawing and embodiment.
With reference to Fig. 1, a system construction drawing of the group buying websites sales data authenticity detection that realizes the application is shown.The application is based on the platform that purchases by group of each group buying websites information of integration and realizes that the sales data authenticity detects.Purchase by group platform through with each group buying websites cooperation, can get access to or analyze obtain each group buying websites purchase by group the cloud data.Group buying websites sales data authenticity detects and then realizes based on purchasing by group the cloud data.At first; Group buying websites sales data authenticity detection system is obtained the single information of current group from the server that purchases by group platform; As a single-character given name claim, describe in detail, sales volume etc. in real time; Again according to a single-character given name claim, information such as detailed description carries out operations such as keyword extraction, from purchase by group the cloud data, determines the single single real-time attention rate of relevant group of current group then, the real-time attention rate that final transformation model will the group's of being correlated with list is converted into theoretical quantity purchase; And compare, thereby the authenticity of the single sales volume of current group is judged with current group single real-time sales volume.Carry out detailed explanation below in conjunction with concrete process.
With reference to Fig. 2, the application's group buying websites sales data authenticity detection method embodiment one is shown, may further comprise the steps:
Step 101 is obtained the real-time sales volume that shows in the single information of current group.
Purchasing by group of a certain single commodity that group singly refers to a certain group buying websites and provided, the real-time sales volume that shows in the single information of current group is obtained through the data interaction between server and the group buying websites server.
Can, the single real-time sales volume of current group obtain when reaching predetermined quantity obtaining of real-time sales volume.If because sales volume is very little in real time, it is very little that false possibility possibly appear in its data, this moment, the judgment data authenticity was also just nonsensical, can increase the unnecessary burden of server on the contrary.Therefore through setting a predetermined quantity, when effective sale quantity reaches or surpass this quantity, obtain again and detect with authenticity.
Step 102 is obtained the single real-time attention rate of current group.
Attention rate is meant the number of times that a list is perhaps browsed by user's normal access, through server user real time is visited or is browsed behavior and monitor and write down and confirm.Concrete, can be through obtaining after the number of times correction that a single-page is clicked.Because; Click to a single-page possibly be that malice is browsed or attacked; In order to guarantee the accuracy of data, need revise the number of times that a single-page is clicked, only keep group's single-page by normal access or browse and the number of clicks that produces as the single attention rate of relevant group.
Concrete, a single-page is discerned and filtered and realize through the ID (User Identification) to the visitor by the correction of number of clicks.Can remove irrational part in the data through this kind processing mode, make the final data that obtains more objective and accurate.
Step 103; Obtain the real-time sales volume that shows in the single information of current group in the same period and the difference of the weighted value of attention rate in real time, if difference, confirms then that the single sales data of current group is untrue greater than threshold value; Otherwise, confirm that then the single sales data of current group is true.
Wherein, the weighted value of attention rate can be real-time attention rate to be carried out normalization handle in real time, makes it to have the normalization processing costs that obtains behind the identical quantized level with real-time sales volume.Because for attention rate and real-time sales volume, the two possibly not belong to identical quantized level, through real-time attention rate is carried out comparing after the weighting again, the result who draws like this is more objective and accurate.This moment can be through in the more same period, and the difference between the changing value of the changing value of sales volume and normalization processing costs judges whether whether sales data is true in real time.If sales data is real words in real time, it should be identical or similar with the variation tendency of attention rate, if the two difference is excessive, explains that then possibly there is false part in sales data.
Attention rate also can be that real-time attention rate is transformed the theoretical quantity purchase that obtains in real time, and the real-time sales volume that shows in the single information of current group this moment is the difference of real-time sales volume and theoretical quantity purchase with the difference of the weighted value of real-time attention rate.Wherein, in real time attention rate be converted into theoretical quantity purchase can be through obtaining to the monitoring of the buying rate of user behavior analysis, existing group forms data with after analyzing in advance.Concrete may further comprise the steps:
S1 obtains the single keyword of relevant group;
S2 confirms the single concrete commodity that purchased by group of relevant group based on the single keyword of relevant group;
S3 confirms the purchase conversion ratio of said concrete commodity;
S4 confirms theoretical quantity purchase according to buying conversion ratio and user's attention rate.
Wherein, buy conversion ratio can based on to various particular commodities sometime the data that purchase by group of section sample definitely, also can purchase by group data through more history and confirm through model training back to particular commodity.Among the application, real-time attention rate is converted into theoretical quantity purchase realizes through predetermined transformation model, following according to the transformation model that abovementioned steps is confirmed:
F-model=a*Log(UF)*TransformRatio(Kw1,Kw2,Kw3...,KwN)
Wherein, UF is the user real time attention rate, and TransformRatio is the purchase conversion ratio function of a certain particular commodity that characterizes of serial keyword, and a is for supplying coefficient.
In order to simplify comparison procedure and data handling procedure; Only if the real-time sales volume of current group list and the difference of theoretical quantity purchase under the more same time point greater than threshold value, confirm that then the single sales data of current group is untrue; Otherwise, confirm that then the single sales data of current group is true.
Preferably; Because sales volume is sometime occurred than great fluctuation process by the various factors influence easily; For example, a lot of users just buy same commodity at one time, and this just possibly cause the sales data of this time point to increase; And make current difference of rolling into a ball single effective sale quantity and theoretical quantity purchase greater than threshold value, so the application can also compare in the following manner:
The single real-time sales volume of current group under each time point in a certain predetermined period and the difference of theoretical quantity purchase add up; If the difference sum that adds up is greater than threshold value; Confirm that then the single sales data of current group is untrue, otherwise, confirm that then the single sales data of current group is true.For example; Predetermined period is 1 hour (10 o'clock to 11 o'clock); Wherein comprise ten time points, can per six minutes once, also can confirm ten time points according to other modes; Calculate the single real-time sales volume and the difference of theoretical quantity purchase in the current group of each time point then, value and predetermined threshold value after the difference of ten time points being added up at last compare again.
Because for the data in certain one-period be affected may much smaller than be affected sometime maybe, whether truly more accurate and objective through the data in certain one-period being judged sales data.
With reference to Fig. 3, the application's group buying websites sales data authenticity detection method embodiment two is shown, in order to guarantee result's accuracy; On the basis of embodiment one; It is single in testing process, to introduce relevant group, and detects according to the single data of relevant group, and detailed process is following:
Step 201, obtain the single information of single group of current group and with group single information in the real-time sales volume that shows.
The single information of group refer to a single-character given name claim, to purchasing by group the information such as description of commodity, roll into a ball single information and in real time sales volume carry out data interaction between can group buying websites server and obtain based on this group's list of server and issue.
Because group's single-character given name is claimed, be changeless to the information such as description that purchase by group commodity; Therefore obtain once and get final product; The real-time sales volume of group's list is along with change of time then possibly change, and therefore the single real-time sales volume of group can be according to preset time, and certain interval of time obtains once.In order to guarantee to obtain the accuracy of data and make things convenient for subsequent analysis, generally, be a few minutes interval time.After obtaining, real-time sales volume and time corresponding are stored, for example, with sequence<Tm (i), S (i)>mode, the real-time sales volume of promptly under the timestamp of certain Tm (i), gathering is S (i).
Step 202 confirms that according to the single information of single group of current group single relevant group of said current group is single.
The relevant Dan Zhiqi of group shows purchase by group commodity and current group single showed purchase by group the identical or similar group's list of commodity.For example, be both the group that purchases by group chafing dish and singly can be confirmed as the group's of being correlated with list, be both and purchase by group the relevant group of also can confirming as of film ticket list or the like, specifically set pattern then can limit based on actual conditions really.
What relevant group was single confirms and can realize through following mode:
Step D1 extracts the single keyword of current group from the single information of single group of current group.
Keyword is the generality word that best embodies the single key property of current group, among the application, keyword from a single-character given name claim and describe in detail extract.Keyword extraction can realize according to common extracting mode, for example, and through keyword extraction based on portmanteau word and synset, or based on the keyword extraction of semanteme.In order to guarantee the single degree of correlation of keyword and current group, extraction that can be the least possible is chosen 3 to 5 keywords generally speaking and is got final product.
Step D2 expands the keyword set that is expanded to the keyword that extracts.
Expansion comprises synonym expansion or notional word amplification expansion.Can obtain more multi-key word through expansion.For accuracy and the correlativity that guarantees keyword, the expansion quantity that the application is general is two times of primary keys quantity.
Step D3 is based on purchase by group cloud data inquiry and the relevant group list definite and expanded keyword coupling of etendue critical set of words from prestoring.
Purchase by group the cloud data and be meant the related data that purchases by group that purchases by group all group buying websites in the platform.For example, merchandise news, keyword, the number of times of being visited or browsing, real-time sales volume, final sales volume that each group of each group buying websites is single, classification under each commodity or the like.Wherein a part of data are directly to carry out data interaction and obtain through purchasing by group interface that Platform Server and group buying websites configure, and group buying websites will purchase by group the data preparation in real time in interface, and server can grasp from interface at any time.Some is the behavioural information that produces purchasing by group on the platform of user (as browse, purchase etc.), and this part data is directly through user behavior monitoring in real time and record are obtained.Purchase by group the cloud data and can and more newly arrive through real-time collection and guarantee ageingly, certainly,, can revise some data through predetermined mode in order to guarantee the accuracy of data, then with revised data also as a part that purchases by group the cloud data.
Keyword through expansion can inquire the group's list that matees with these keywords from purchase by group the cloud data, these groups singly are the single relevant group list of current group.
Step 203 is confirmed in real time total attention rate that single and relevant group of current group is single.
Always attention rate is meant the total degree that the Dan Tuandan of single and relevant group of current group is perhaps browsed by user's normal access, monitors and writes down and confirm through user real time is visited or browsed behavior.Concrete, can be through obtaining after the total degree correction that a single-page is clicked.Because; Click to a single-page possibly be that malice is browsed or attacked; In order to guarantee the accuracy of data, need revise the total degree that a single-page is clicked, only keep group's single-page by normal access or browse and the number of clicks that produces as the single attention rate of relevant group.
Concrete, a single-page is discerned and filtered and realize through the ID (User Identification) to the visitor by the correction of number of clicks.Can remove irrational part in the data through this kind processing mode, make the final data that obtains more objective and accurate.
Step 204; Obtain the difference of the weighted value of single and relevant the single in real time total attention rate of real-time sales volume and said current group that shows in the single information of current group in the same period; If difference is greater than threshold value; Confirm that then the single sales data of current group is untrue, otherwise, confirm that then the single sales data of current group is true.
Be appreciated that; The difference of the weighted value of single and relevant single in real time total attention rate of real-time sales volume that shows in the single information of current group in the same here period and said current group is identical with definite method of the difference of the weighted value of real-time attention rate with the real-time sales volume in the previous embodiment one; Can confirm with reference to aforesaid way, repeat no more at this.
With reference to Fig. 4, a kind of group buying websites sales data authenticity pick-up unit embodiment one of the application is shown, comprise the single information acquisition module of current group 10, attention rate determination module 30 and comparison module 50 in real time.
The single information acquisition module 10 of current group is used for obtaining the real-time sales volume that the single information of current group shows.
Attention rate determination module 30 is used to obtain the single real-time attention rate of current group in real time.
Preferably, in real time attention rate determination module 30 comprise number of clicks acquiring unit, amending unit and in real time attention rate confirm the unit.Wherein, the number of clicks acquiring unit is used to the number of times that the group's of obtaining single-page is clicked.Amending unit is used for the said number of times correction of being clicked is obtained the single number of times of perhaps being browsed by user's normal access of said group.The attention rate determining unit is used for confirming that by the number of times that user's normal access is perhaps browsed be the single real-time attention rate of current group in real time.Further, amending unit also comprises ID recognin unit, is used for visitor's ID is discerned and filtered.
Comparison module 50; Be used for the single information of current group shows in the more same period the real-time sales volume and the difference of the weighted value of attention rate in real time,, confirm that then the single sales data of current group is untrue if greater than threshold value; Otherwise, confirm that then the single sales data of current group is true.
Preferably, this device also comprises trigger module, is used for when the real-time sales volume that the single information of current group shows reaches predetermined quantity, triggering the real-time sales volume of obtaining current group list.
Preferably; This device also comprises conversion module 40 (as shown in Figure 5); When the real-time sales volume that shows in the single information of current group when in real time the difference of the weighted value of attention rate is the difference of real-time sales volume and theoretical quantity purchase, conversion module 40 is used for real-time attention rate is converted into theoretical quantity purchase.Further, conversion module 40 comprises that keyword acquiring unit, commodity confirm that unit, conversion ratio confirm unit and transformation model.Wherein, the keyword acquiring unit is used to obtain the single keyword of current group.Commodity are confirmed the unit, are used for confirming the single concrete commodity that purchased by group of current group based on the single keyword of current group.Conversion ratio is confirmed the unit, is used for confirming the purchase conversion ratio of said concrete commodity.Transformation model is used for confirming theoretical quantity purchase according to buying conversion ratio with real-time attention rate.
With reference to Fig. 5, the application's group buying websites sales data authenticity pick-up unit embodiment two is shown, comprise the single determination module of the single information acquisition module of current group 10, relevant group 20, attention rate determination module 30, conversion module 40 and comparison module 50 in real time.
The single information acquisition module 10 of current group is used to obtain single information of single group of current group and real-time sales volume.
The single determination module 20 of relevant group is used for confirming that based on the single information of single group of current group single relevant group of said current group is single.
Preferably, the single determination module 20 of relevant group comprises keyword extracting unit, expanding element and query unit.Wherein, keyword extracting unit is used for extracting the single keyword of current group from the single information of single group of current group.Concrete, extract current single keyword of rolling into a ball single title and the detailed description from current the group.Expanding element is used for the keyword set that is expanded expanded in the keyword that extracts.Query unit is used for purchasing by group the inquiry of cloud data and confirming single with the relevant group of expanded keyword coupling from what prestore based on the etendue critical set of words.Attention rate determination module 30 is used to obtain the single in real time total attention rate of single and relevant group of current group in real time.
Conversion module 40 is used for real-time attention rate is converted into theoretical quantity purchase.What at this moment, conversion module 40 transformed is the single in real time total attention rate of single and relevant group of current group.Its process is identical with the real-time attention rate that transforms current group list, repeats no more at this.
Comparison module 50; Be used for obtaining the difference of the weighted value of single and relevant single in real time total attention rate of real-time sales volume and said current group that the single information of current group shows in the same period; If difference is greater than threshold value; Confirm that then the single sales data of current group is untrue, otherwise, confirm that then the single sales data of current group is true.
Each embodiment in this instructions all adopts the mode of going forward one by one to describe, and what each embodiment stressed all is and the difference of other embodiment that identical similar part is mutually referring to getting final product between each embodiment.For device embodiment, because it is similar basically with method embodiment, so description is fairly simple, relevant part gets final product referring to the part explanation of method embodiment.
More than group buying websites sales data authenticity detection method and device that the application provided are described in detail; Used specific case herein the application's principle and embodiment are set forth, the explanation of above embodiment just is used to help to understand the application's method and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to the application's thought, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as the restriction to the application.